基于互信息和K-means聚类的信息安全风险评估

Risk Assessment of Information Security Based on Mutual Information and K-means

针对信息安全风险评估训练数据少、求解最优值困难等问题,提出了基于互信息和K-means聚类的信息安全风险评估方法.用模糊评价法量化风险指标,通过互信息计算风险因素与风险等级之间的依赖性,找出风险度在每个等级的最优点作为K-means初始中心点,用K-means算法对数据分类.该方法实现简单且克服了K-means对初始值敏感和现有的信息安全风险评估方法求解最优值困难、结论模糊等缺陷.实验结果证明了其有效性.

This paper proposes a new information security risk assessment method based on mutual information calculation and K-means clustering algorithm in order,to solve the problems of small training data and to find optimal value.Fuzzy evaluation is applied to quantify risk factors,calculating the mutual information value of risk factors to indicate the dependent degree of risk factors and risk degree.The data are classified by K-means with the optimal mutual information data as the initial centers.This method is less computation,and also can overcome the K-means's shortcoming sensitive to initial value and solve problems of optimal value and conclusion blurring defects.Experimental results show the effectiveness of the method.